Contact Area and its Relation to Friction?

[Joachim]

"Recently we did a tire traction experiment . . ."

Loolking at your results, which were for dynamic friction on carpet, the highest effective coefficients of friction were achieved by the IFI wheels, and with the IFI wheels, the highest coefficients of friction were achieved by the wheels with the most lightly loaded contact patches.

In other words, in your tests, for the IFI wheels, for a given weight, more contact area gave more friction. This is a fairly common result for high-traction materials on smooth surfaces. Looks like it could apply on carpet too.

[Lil' Lavery]

Originally Posted by Joachim

"Recently we did a tire traction experiment . . ."

Loolking at your results, which were for dynamic friction on carpet, the highest effective coefficients of friction were achieved by the IFI wheels, and with the IFI wheels, the highest coefficients of friction were achieved by the wheels with the most lightly loaded contact patches.

In other words, in your tests, for the IFI wheels, for a given weight, more contact area gave more friction. This is a fairly common result for high-traction materials on smooth surfaces. Looks like it could apply on carpet too.

Those results, as you so kindly bolded, were dynamic friction, not static friction. I don't believe anyone here contested more contact area equating to more dynamic friction. The center of the debate, however, was on the relation between contact area and static friction.

[Alan Anderson]

Originally Posted by the spreadsheet

The force is the dynamic force to keep the tire sliding on the carpet once it is started.

Dynamic friction does depend on contact area.

Since static friction is greater than dynamic friction, your results don't say anything about what wheels would win a pushing contest. A robot with the wheels not sliding on the carpet is generally going to beat out a robot that's spinning its wheels.

[Joachim]

Those results . . . were dynamic friction, not static friction . . . .

Yes, but the two are often, though not always, correlated. The standard rule (or standard model) for both static and dynamic friction (Amonton's or Amonton/Coulomb laws) is that that friction is the product of the load and the appropriate static or dynamic coefficient for the materials in contact, and is independent of the contact area. See, for example, http://hyperphysics.phy-astr.gsu.edu...frict.html#fri (Article on friction at the Hyperphysics site of Georgia State). Thus an experiment showing that the dynamic coefficient of friction varies with contact area might lead one to at least adopt the hypothesis that the coefficient of static friction does the same.

The truth is that the nice linear model for both static and dynamic friction taught in basic physics and engineering courses only works well for some materials under some conditions. Rubber is not one of those materials. See, for example, http://www.springerlink.com/content/n30715161g635138/ (abstract of a book chapter entitled "The Influence of Contact Pressure on the Dynamic Friction Coefficient in Cylindrical Rubber-Metal Contact Geometries") which states in part: As it is commonly know[n], classic Coulomb’s and Amonton’s friction laws, which mainly establish that the friction coefficient is independent of the area of contact, are proven to be not valid in the case of rubber-like materials. See also http://www3.interscience.wiley.com/c...69929/ABSTRACT (abstract of an article entitled "Analytical and experimental investigation of the static friction regime for rubber-rigid ball contact") which states in part: The parameters of the static friction regime in terms of static friction force . . . are investigated for a rubber ball/metal flat configuration. . . . The coefficient of static friction decreases significantly by increasing the normal load . . . . Smaller radii of the ball determine a smaller static friction force . . . .

Where rubber is involved, increased contact area (wider tires, larger radius tires, or belts/treads) and lower contact loads often have the effect of increasing the available traction (friction), whether static or dynamic.

[cobrawanabe1699]

Alrighty. This is where being a gearhead my entire life pays off.

FACT 1:
As a vehicle turns, the contact patch (and weight) move to the outside of the tires. re: if a vehicle makes a right turn, the contact patch moves to the left sides of the tires. That's why you see road racing cars with lots of negative camber (wheels angled inward towards the top). This keeps the contact patch closer to the insides of a tire, and when the car makes a corner, the outside tire keeps more contact with the ground.

FACT 2:
As a vehicle accelerates, weight tranfers to the rear, making the rear contact patch wider, and the front contact patch thinner. That's why FWD cars STINK at drag racing.

FACT 3:
The faster a vehicle is moving, the more it's weight tranfers in a corner. That's why cars get "body lean", or leans towards the outside of the corner. Some cars (especially Volkswagen Corrados) even lift the inside rear tire when braking and turning hard.

FACT 4:
THE MORE CONTACT THE BETTER, unless on a soft, malleable surface, such as snow.

FACT 5:
The stiffer the chassis, the more contact all tires will have with the ground at all times. Hence, a flimbsy chassis will handle MUCH worse than one that stays stiff and keeps all tires in contact with the ground.

FACT 6:
The lower the center of gravity, the better. That's why cars lowered on a good suspension setup (stiff) always handle better than stock. The higher the center of gravity the vehicle has, the more the weight will transfer. Say a vehicle is making another right-hand turn. A high center of gravity will lead to very thin contact patches on the outsides of the right-side tires, and a contact patch that may even shrink towards the outsides of the left-side tires.



These are simple terms. I am taking my first physics class, and I source all of this information from my vehicluar knowledge. PLEASE CORRECT ME IF THESE FACT SEEM INNACURATE. I see no need to break the traction of a 10 ft/s robot down into static or dyanic friction. You simply need to know where that friction is and how to maximize it.

[Woody1458]

Originally Posted by squirrel

Friction is weird, so yeah, you probably do have to derive a formula from empirical data.

An example of friction being weird: Most material interfaces have a higher coefficient of static friction, than of dynamic friction. But aluminum to aluminum has a higher coefficient of dynamic friction than static friction.

And when you consider automobile tires, think about what might be happening when the situation in my avatar occurs....

Good luck!

Although I'm only a high school student I must question your comment on aluminum on aluminum friction. If the static friction is lower then kinetic, this means as soo as you start movement with minimum force it will stop going back to static then starting again then moving imediatly back to static. Thus never moving at all, thus always staying as static friction. Am I missing something?

Also as an answer to the question mu the mathmatical ratio of the force of friction/normal force is unique not only for every material but for every object. Thus, the contact area is calculated in when stating mu. I couldn't give you a mu of rubber on asphalt, but i could give you a mu of a specific tire with a specific contact area and texture.

just as an explination of your source, I take Honors Physics and got a 20/20 on my friction quiz (no tests in this class) =P

[lukevanoort]

Originally Posted by Woody1458

Also as an answer to the question mu the mathmatical ratio of the force of friction/normal force is unique not only for every material but for every object. Thus, the contact area is calculated in when stating mu. I couldn't give you a mu of rubber on asphalt, but i could give you a mu of a specific tire with a specific contact area and texture.

Sort of. In classical friction you ignore the effect of surface area and the mu doesn't take it into account, which provides a very good approximation with hard materials. On the other hand, with interlocking materials (which I am still not convinced applies to FIRST carpet), there is something of an effect from surface area, but it is probably (I have never tried) a bit hard to predict well enough to warrant including in calculations. While the best results for traction tests would be between the same two objects (manufacturing irregularities would get taken into account) and testing them, similar objects can work pretty well. For example, if I wanted a very accurate number for how much traction a robot has, I would take it to a competition field with a force gauge. Unfortunately, that is unrealistic, so one can just test with a sample of the material on similar (ideally the same) carpet, and still get pretty decent numbers.